cd ~/

Defining V2: Learning from V1 in the Field

Translating field feedback and technical learnings into requirements for the next-generation device.

Evyatar Bluzer
2 min read

V2 planning is underway. The question isn't "what do we want?" but "what did V1 teach us we need?"

V1 Pain Points (Perception)

From telemetry and user feedback:

Tracking failures in challenging lighting: 35% of reported issues

  • Users want to use device near windows
  • Evening use in dim rooms
  • Transitions (walking between rooms)

Relocalization unreliability: 12% failure rate

  • Maps become stale
  • Users don't understand map management
  • Content "jumps" unexpectedly

Hand tracking limitations: Frequent complaint

  • Gesture recognition was cut
  • Tracking range too limited
  • Performance degrades with occlusion

Battery life: Sessions limited by power

  • Heavy use drains in 2.5 hours
  • Thermal throttling kicks in sooner

V2 Perception Requirements

Tracking Robustness

  • Target: under 1 tracking loss per hour of normal use (10x improvement)
  • Approach: Better sensors + smarter algorithms + wider operating envelope

Sensor Upgrades Needed

V1V2 TargetWhy
Tracking cam: 640x4801280x720+More features, better low-light
Depth: 320x240 ToF640x480Better meshing, hand tracking
Eye cam: 320x320480x480+More accurate gaze
IMU: 1kHz2kHz+Better fast motion handling

Hand Tracking

  • Full gesture recognition (pinch, grab, point, palm)
  • Near-range: 20-60cm from face
  • Far-range: 40-100cm (arm's length)
  • under 5° fingertip accuracy

Power Budget

  • Target: Same or better battery life with more capability
  • Approach: New chip architecture, algorithmic efficiency, smart power management

Architecture Decisions

Dedicated ML accelerator: V1 ran ML on GPU. V2 needs NPU for efficiency.

More on-device compute: Cloud isn't reliable enough for real-time. Process locally, sync when possible.

Sensor-algorithm co-design: Don't pick sensors then adapt algorithms. Design together.

What We're Not Changing

Some V1 decisions remain correct:

  • Visual-inertial approach for tracking (proven, robust)
  • ToF for depth (better outdoor than structured light)
  • Stereo tracking cameras (needed for stereo depth backup)

Timeline Pressure

Chip tape-out is in 6 months. Sensor orders locked in 4 months.

We need requirements finalized NOW for hardware that ships in 2+ years.

The V1 experience helps - we know what matters. But we'll inevitably get some things wrong. Building in flexibility where we can.

Comments